The perceptual basis of spatial sound perception

Our ability to derive spatial impressions from a sound field is based on the facts that we have two sensors which are spatially separated by typically 18 cm and that the space in between these sensors is filled by acoustically nontransparant material. The first fact leads to a time difference at the two ears for sound coming from a lateral direction, which increases monotonically from 0 (frontal source) to about 0.7 ms (source at about 90 degrees). The second fact leads to interaural level differences which, on average, increase with increasing laterality and which are, on average, larger at high than at low frequencies. A third contribution to spatial impression comes from the direction-specific filtering through the outer ear (pinna) which is highly characteristic for each individual and which is the basis for deriving the elevation of a sound source. Besides cues allowing the determination of direction, sound fields can also comprise cues allowing the determination of the distance of a sound. These cues are much richer in enclosures, were the ratio between direct and reflected sound energy can be exploited by human listeners. In a free field, acoustic distance cues are the overall intensity decrease with increasing distance and the increased attenuation at high frequencies due to air absorption. These cues can, however, only be used if the listener has some knowledge about the spectrum and level of the source signal. Finally, some examples for the influence of visual cues on human spatial sound perception will be described